Open Access
Translator Disclaimer
September, 1978 Linear Estimation for Approximately Linear Models
Jerome Sacks, Donald Ylvisaker
Ann. Statist. 6(5): 1122-1137 (September, 1978). DOI: 10.1214/aos/1176344315

Abstract

An approximate linear model is proposed to allow for deviations from an underlying ideal linear model as follows: If, in standard notation, $Y = A\beta + \varepsilon$ is the ideal model then $Y = A\beta + r + \varepsilon$ where $|r_i| \leqq M_i$ for $M$ a given vector is an approximate linear model. The problem solved here is that of finding a linear estimate of a single linear function of $\beta$ which minimaxes mean square error in the approximate model. The estimate obtained may be the standard one from the ideal model, but in general it is not. The estimate is calculated as a solution to a set of nonlinear equations (generalizing the usual normal equations) and an algorithm is given for obtaining the solution.

Citation

Download Citation

Jerome Sacks. Donald Ylvisaker. "Linear Estimation for Approximately Linear Models." Ann. Statist. 6 (5) 1122 - 1137, September, 1978. https://doi.org/10.1214/aos/1176344315

Information

Published: September, 1978
First available in Project Euclid: 12 April 2007

zbMATH: 0384.62058
MathSciNet: MR501615
Digital Object Identifier: 10.1214/aos/1176344315

Subjects:
Primary: 62J05
Secondary: 62J10 , 62J35

Keywords: approximately linear models , minimum mean square linear estimation , normal equations

Rights: Copyright © 1978 Institute of Mathematical Statistics

JOURNAL ARTICLE
16 PAGES


SHARE
Vol.6 • No. 5 • September, 1978
Back to Top